cim tv - methodologie · cim tv: calculation and reporting rules – january 2016 3 introduction...

CIM Television

Calculation and Reporting Rules

January 2016

CIM TV: Calculation and Reporting Rules – January 2016 2

These exploitation rules were created with the much appreciated assistance of:

- SKO from the Netherlands who gave their permission to use elements and formulas from the document ‘Calculation and Reporting Rules’,

- Philippe Delhaise (Mediaxim) who compiled the basic copy regarding formulas and who corrected various revisions with the patience of angels.


Introduction

The CIM has entrusted the CIM TV study for the period 2016-2017 to GfK Belgium for the

audience study.

The contract for the time logging has been entrusted to Nielsen for the period of 2013-2016.

The aim of this document is to indicate how data should be calculated and how it can be

used.

As part of the introduction of time shifted viewing, CIM has certified all TV software

applications of the so-called TV software houses.

A list of certified companies is provided at www.cim.be.


Table of Contents

Introduction ............................................................................................................... 3 1. Usage of the data ................................................................................................... 5

1.1 Rules concerning the use of CIM TV data .............................................................. 6 2. Calculation rules ..................................................................................................... 8

2.1 Calculation rules “live” ........................................................................................ 8 2.2 Calculation rules “live+7” .................................................................................. 18 2.3 Algorithm for the evaluation of advertising campaigns (“Live+7”) ........................... 25 2.4 Calculation rules for “Live+Playback” .................................................................. 32

3. Reporting rules ..................................................................................................... 35 3.1 Live, Live+7 and Live+Playback rating/reach .......................................................... 35

3.2 Live, Live+7 and Live+Playback ......................................................................... 36 3.3 Source citation ................................................................................................. 39 3.4 Rounding ........................................................................................................ 39

4. Available files ....................................................................................................... 40 4.3 Top 20 and Top 100 ......................................................................................... 40


1. Usage of the data

The CIM TV data are available to all subscribers of the CIM TV study. A list of entitled

subscribers is posted at http://www.cim.be/fr/television/liste-des-souscripteurs

The subscribers must respect the rules regarding the communication of data. These rules

are available at www.cim.be and in paragraph 3.

Only CIM-certified software houses may warrant use of CIM TV data. They are held to strict

confidentiality and may under no circumstances use the CIM TV data in any other form.

Every situation that is cause for the slightest doubt should be submitted to the CIM.

The issue of these calculation and reporting rules is prompted by the availability of data

regarding time shifted viewing in the Belgian market.

Until then, the currency in TV land is the familiar ‘live’ viewing on the day of

broadcast.

After the introduction of time shifted viewing, this has changed: the currency was all

of ‘live’ television watching + all time shifted viewing (the definition follows in 2.2.1).

In 2013 a new currency was introduced into the study: Live+6+Guests: all of ‘live’

television watching + all time shifted viewing + viewing behaviour of guests.

In 2016 a new currency was introduced into the study: Live+7+Guests. The time

shifted viewing was extended until the 7th day after broadcasting.

Analysis and reporting on the basis of other viewing definitions remain possible, but they

should be clearly specified and can no longer serve as currency.

http://www.cim.be/


1.1 Rules concerning the use of CIM TV data

1.1.1 Data for Subscribers

Two files are generated on a daily basis for subscribers of the CIM TV study:

ATA file: data per second, complete list of all TV channels.

These files are made available to TV channels and their national advertising producers,

rank 1 subscribers, who have subscribed for this.

ATB file: data per second for national subscribed TV channels in addition to aggregated

data per quarter for all other TV channels who earn at least 1% market share in the

complete universe (4 years and older) (only ratings and share).

This file thus contains 2 different levels of detail depending on the channel, and the

number of reported channels is limited. These files are made available to intermediaries.

The following regroupings are included:

Other channels – Dutch-speaking

Other channels – French-speaking

Other channels – neither Dutch nor French nor satellite, VCR + DVD+Perif., Pay TV +

DVR, Play console + other screen usage

1.1.2 Data for other CIM members

CIM members who are not subscribed to the CIM TV study can request monthly channel

reports (MZR). A report can be provided by a choice of target group and hour slot with the

average day reach, week reach, month reach and market share of a TV channel.

1.1.3 Which data can be included in external communication?

The data that every subscriber of the CIM TV study can use in its external communications

are only the data for channels that are found in the ATB file, the pre-calculated channels or

channels with a monthly broadcasting reports (MZR). In other words, communicating the

data found in the ATA file externally that are not found in the mentioned files above is

prohibited.

It is also precluded for the CIM that a TV channel is compared on the basis of figures per

second with a TV channel on the basis of quarterly figures. If one of the TV channels is

reported per quarter in the AUB file, in a comparison with other TV channels, the latter must

also be indicated on the basis of data per quarter, even if data per second is available.

In the published figures, one must draw a distinction between CIM TV figures and estimates

based on CIM TV figures. In the latter case, one must state ‘estimate of ... based on CIM TV

figures’.

The terminology and calculation rules as indicated in the annual report of 2000-2001* must

be respected by the users of the CIM TV data. * Can be downloaded from http://cim.be/media/Televisie

http://cim.be/media/Televisie


From 1 January 2006, the variable of Total TV was redefined and is now in line with the

European standard as used by most countries.

Total TV contains in its new definition of the following elements:

All individual broadcasters (registered via EAM, Picture Matching and/or CNI code)

‘Other TV’ broadcasters

Satellite Receiver

In addition, there is the variable Total screen usage.

As the term suggests, this regards the total number of all screen use. It is mainly the sum of

Total TV plus the following ancillary equipment:

Video recorder

Other hardware device (camera, PC, etc.)

Game console (Sony PlayStation, Microsoft X-Box, etc.)

Set-top box (Telenet Digibox, Proximus TV, etc.)

Pay-TV decoder (Prime, Canal+, BE TV, etc.)

DVD (reader, writer, reader + VCR, etc.)

Connected TV (Smart TV’s, Chromecast, Apple TV,…)

Minimum number of individuals in the choice of target groups

The following guideline also applies: a minimum of 300 individuals must be represented

in the random sample size of a selected target group.

If the number of 300 individuals is not achieved, the CIM requests that this be mentioned in

the reporting.

For more information, please contact CIM:

02 661 31 53 ([email protected])

mailto:[email protected]


2. Calculation rules

2.1 Calculation rules “live”

2.1.1 Reference universes

The following definitions of reference universes have already been in use since 1997

(Methodological report CIM TV – version of March 2007) and do not change. They are

included here for the sake of completeness.

The reference universe is the population for which the study wants to establish a

representative measure of television viewing behaviour. The definition of the reference

universe remains constant but numbers vary according to the calendar year, depending on

how the population evolves.

The Total Reference Universe is the Dutch-speaking and French-speaking population aged

4 and over, residing on Belgian territory (including foreign nationals), and who belong to a

household that has at least 1 television set at its disposal.

Because of the regional division of Belgium, in practice two sub-universes have been

delineated.

North Universe: all people who meet the above definition and that live in the region of

Flanders or Brussels (defined as the Nielsen 3 region) and who are Dutch-speaking.

South Universe: all people who meet the above definition and that live in the region of

Wallonia or Brussels (defined as the Nielsen 3 region) and who are French-speaking.

The composition of the television panel may vary slightly per day: sometimes new people

are added, other people leave the panel, sometimes the information for a family is missing

for technical reasons... To obtain representative viewing figures every day, all members of

the panel are thus weighted every day to obtain a perfect match with the reference

universe. This ‘daily weight’ is a basic parameter for the correct calculation of viewing

figures.


2.1.2 Broadcast parameters

When identifying television broadcasts, different parameters are used by the CIM for time

logging:

Type: A = programme, B = break, C = auto-promotion.

Genre: e.g. 11=film.

Title: description of the time-logged element.

Level: level 1, 2 or 3:

The TV-log is in fact like a tree. Some programs of level 1 can contain elements of

level 2 (program, break or auto-promotion). Programs of level 2 can also contain

elements at level 3.

Section: if an element can be considered as a section of a ‘father’ program, the value

is 1; in all other cases it is 0.

A classical example are parts of a film or series separated by breaks and/or auto-

promotions. Another example would be the different parts of a football program.

You can find the list of Genre codes on the methodology page of the CIM TV section of the

CIM website.

http://www.cim.be/sites/default/files/Media/Televisie/Documents/CIM%20TV%20-

%20Horodatage%20-%20New%20typology%20-%20Final%20-%20List_0.pdf

These parameters are included in the results of the CIM TV study. The following chart

provides an illustration:

A ‘BRUTO’ (gross) programme is defined by a start and an end of a programme, without

extracting parts as commercial spots, auto-promotion, breaks,...

A ‘NETTO’ (net) programme is defined by a start and an end of a programme, excluding the

commercial spots, auto-promotion,... = netto programme.

http://www.cim.be/sites/default/files/Media/Televisie/Documents/CIM%20TV%20-%20Horodatage%20-%20New%20typology%20-%20Final%20-%20List_0.pdf

http://www.cim.be/sites/default/files/Media/Televisie/Documents/CIM%20TV%20-%20Horodatage%20-%20New%20typology%20-%20Final%20-%20List_0.pdf


2.1.3 The CIM audience Threshold

a. Viewing duration

The sample data is available per second. For some metrics, the CIM has set a minimum of

viewing time.

For example viewing for at least 1 min / 10 min =

Anyone who watched a programme, a time band or a channel at least 60/600

seconds.

This needs regrouping of different STATS-lines per viewing type (live, TSV):

HOME-ID

MEMB-ID or GUESTID

TV-NUMBER

CHANNEL or VIRTUAL CHANNEL

ACTVITY CODE = TV and Overlay

This regrouping is sufficient for non-consecutive viewing.

If the viewing needs to be consecutive the regrouping is more stringent. The conditions are

enlarged to identify consecutive viewing of a person in case of different statements (e.g.

different statements due to arrival of other panel members or guests):

The regrouping of different STATS-lines per viewing type:

HOME-ID

MEMB-ID or GUESTID

TV-NUMBER

CHANNEL or VIRTUAL CHANNEL

ACTVITY CODE = TV and Overlay

END TIME STATSx-1 = STARTING TIME STATSx

As the lower threshold of a programme reach, the CIM has set a minimum of 60 consecutive

seconds viewing a programme.

In addition, the user is free to determine or set a higher minimum viewing duration and

whether this must be consecutive or not, and to choose whether someone must have

watched one or more programmes to count as part of the programme reach.

How long someone must have watched one or more programmes can either be stated in

minutes or as a percentage.

Frequently occurring values include: a minimum of 1 consecutive minute, a minimum of 10

minutes and a minimum of 10 consecutive minutes.

b. % of viewing

The CIM defines viewing a spot as having contact with a spot during at least half of its

duration in its time slice. See 2.1.4 e.


2.1.4 Rating calculation rules

a. Programmes

The rating of a programme is defined as the average percentage of viewers per second

within the reference universe that watched this programme. The respondents in the random

sample are weighted with a weight determined by the extent to which they occur in the total

population of viewers in the reference universe.

%100

1

1

,

R

r

r

R

r

irr

i

w

pw

rating

ratingi :rating for programme i

R: number of respondents in the sample

wr: weight of respondent r

p r,i: proportion of the programme i viewed by respondent r

Where the proportion is equivalent to:

i

ir,

ir,duration

p

V

V r,i: Viewing duration of respondent r for programme i

Duration i: duration of programme i

b. Time bands

The rating of a time band is defined as the average percentage of viewers per second within

the reference universe that watched this/these channel(s) during this time band.

%100

1

1

,

R

r

r

R

r

trr

t

w

pw

rating

ratingt : rating for time band t



p r,t: proportion of time band t viewed by respondent r

Averages per time band, over multiple days, should be calculated as follows: a data file is

calculated on a daily basis for the relevant target group including the rating per channel, per

second, rounded up to a minimum precision of six decimals. These data are then used to

calculate averages over multiple days, rounded off to the desired number of decimals.


c. Total TV

The Total TV rating is defined as the average percentage of viewers within the reference

universe who watched one channel or another during a programme or time band.

Z

z

t,zt,Z ratingrating1

ratingZ,t: Total TV rating for time band t

rating,z,t : rating for channel z in time band t

Z: total number of channels

Total TV comprises the following elements:

All individual channels (registered via EAM and/or Picture Matching and/or CNI

code)

‘Other TV’ channels

Satellite Receiver

d. Total Screen Usage

As the term suggests, this regards the total number of all screen use. It is mainly the sum of

Total TV plus the following ancillary equipment:

Video recorder

Other hardware device (camera, PC, etc.)

Game console (Sony PlayStation, Microsoft X-Box, etc.)

Set-top box (Telenet Digibox, Belgacom TV, etc.)

Pay-TV decoder (Prime, Canal+, BE TV, etc.)

DVD (reader, writer, reader + VCR, etc.)

BEWARE: both Total TV and Total Screen Usage should be based on the Station code (never

the Source code).1

e. Spots

Viewing a spot is a Boolean variable: it is true (1) or false (0). The CIM defines viewing a

spot as having contact with a spot during at least half of its duration in its time slice.

Timings and durations are expressed in integer numbers and can be even or odd

Let us define:

l(s) = the length of a spot s in seconds

nS(r,s) = the number of seconds of spots viewed by respondent r, consecutive or

not.

1 The Station code (field 2) indicates which channel or station has been watched. This might be an individual channel (registered via EAM and/or Picture Matching and/or CNI code) or any other peripheral equipment (DVD, DVB, VCR,..). This value is different from the Source code (field 9). The Source code indicates which device generates the signal: 0=TV, 1=VCR, 2=SATELLITE, 3=PAY, 4=DVD, 5=DVB, 6=PLAY, 7=OTHER DEVICE, 8=Connected TV (see 5.3 Description of ATA-ATB and TSV files).


The contact value of spot s for respondent r is then a variable with two states:

c(r,s)=1 (true), if nS(r,s) >= ceiling(l(s)/2)

c(r,s)=0 (false), if nS(r,s) <= ceiling(l(s)/2)

Since time is an odd or even integer, the ceiling function is used to transform the result of

the division in the nearest integer.

Example for a spot of 21 seconds:

c(r,s) =1 if nS(r,s) >= 11

c(r,s) =0 if nS(r,s) < 11

The rating of a spot is defined as the percentage of viewers within the reference universe

who have seen at least half of the spot.

%100

(

1

1

),

R

r

r

R

r

srr

s

w

cw

rating

ratingt: rating for spot s



pr,s: =1 if respondent r saw nS(r,s) or more seconds of this spot with NsecondsMin=

ceiling(lengths/2), else = 0.

f. Blocks of advertising

A block of advertising usually starts and ends with a jingle. The rating of the block is not

calculated on the basis of the individual ratings of the spots, but on the basis of the starting

and ending time of the block. The jingles are included in the calculation of the rating of the

blocks of advertising.

The calculation is identical to that of programmes and time bands (see a. and b. above).

2.1.5 Market share calculation rules

Until the introduction of Time Shifted Viewing, the market share was defined as the

percentage of viewers of a broadcast or time band, based on the total viewing audience.

Market share is calculated by dividing the rating of a programme or time band by the rating

of total channels in the relevant programme or time band.

t,Z

tz,

tz

rating

ratingemarketshar ,

Both Total TV and Total Screen Usage can be used as a reference within the software

application.


2.1.6 Reach calculation rules

a. Programme

The programme reach is defined as the percentage of viewers who watch one programme at

least once for a certain time.

%

w

dw

reachR

r

r

R

r

I,rr

I 100

1

1

Jj

Jjd

Ir

IrIr

,

,,

0

1

I : series of programmes

dr,i: ‘dummy’ that indicates whether respondent falls within the programme rating according

to specific definition

J: criterion for the specific programme rating (e.g. 60 consecutive seconds)



jr,I: calculated according to the specific conditions, personal criterion of respondent for the

programme series I

b. Reach of a channel

The reach of a channel is defined as the percentage of viewers who have watched a certain

channel at least once for a certain time (duration).

2.1.7 Viewing Time calculation rules

The viewing time is defined as the average time that viewers watched a programme or a

channel during a time band, expressed in seconds.

R

r

r

R

r

trr

t

w

Vw

time viewing

1

1

,

viewing time t: viewing time in time band t

R: number of respondents in the random sample


Vr,t: viewing time of respondent r in time band t


2.1.8 Special calculation rules

a. Genre average

As indicated above (see 2.1.2) there is a tree structure in the genre codes of TV broadcasts.

When determining the average reach of genre codes, the fact that the same genre can occur

at different levels must be taken into consideration. This creates the risk of a broadcast

being counted twice.

The simplest way to avoid this is to only calculate the average for identical levels: 1, 2 or 3.

Combining different codes is only permitted if it is unquestionably clear that no minutes are

counted double.

As an example we take the next scheme from paragraph 2.1.2 of one football match:

If one wants to calculate the average for code faa for different football matches, this must

be done at the same level: the average of every part of the game (level 3) or the average of

level 2 or 1. Attention: fal stands for interviews before and after the game.


b. Multiple broadcasts

Cumulative rating

The cumulative rating is defined as the added ratings for multiple broadcasts.

Cumulative rating: cumulative rating for n programmes

n: number of programmes

ratingi: rating for programme i

Weighted average rating

The weighted average rating is defined as:

Weighted market share

The weighted market share as being

c. On the basis of daily sampling

Daily sampling is used for all results that pertain to one day and to averages over multiple

days.

A daily sample includes all panel members of families who were successfully polled for that

day. This involves all individuals for which the data is contained in the demographic files of

gross files. No a posteriori adaptations or corrections are to be made.

d. On the basis of a periodic sampling (more than 1 day)

The weighted factor for each respondent from the periodic sampling is calculated by taking

the mean of the daily weights of the respondent. The denominator is the total number of

days of the periodic sampling on which a broadcast occurred.

When dealing with target groups over a longer period of time, some Panellists may join or

leave this target group. A panellist could for example have his 30st birthday within the

analysed period, and no longer be part of the target group ‘Females 18-29 yr’. If socio-

demographical characteristics change within the analysed period, only the days on which the

panellist has the characteristics of the target group should be taken into account. All other

days he should receive a weight of 0.

e. Guests

The age and gender of each registered guest is known (see 4.1). Until 1/1/2013 guests were

not included in the calculations.


Guests were only taken into consideration for the weekly top 20 and annual top 100 of the

CIM TV figures.

From 1/1/2013 guests are profiled in the CIM TV study and guest viewing is part of the

official currency of the CIM TV study. The viewing behaviour of the guest is linked to a panel

member that isn’t watching at that time. The link based on 4 parameters: region (North –

South), gender, age category (4-11, 12-14, 15-24, 25-34, 35-44, 45-54, 55-64, 65+) and

type of set top box.

The documentation on guest viewing is available on the CIM website:


%20Info%20guest%20viewing_FR.pdf.


%20Info%20guest%20viewing_FR.pdf

f. Rounding

All calculations take place without rounding.

If the result types are determined in the meantime, they are rounded to a minimum of 6

decimal places. The number of possible deviations in respect of the reference is thus

negligible.

http://www.cim.be/sites/default/files/Media/Televisie/Documents/CIM%20TV%20-%20Info%20guest%20viewing_FR.pdf





2.2 Calculation rules “live+7”

2.2.1 Basic Principle

Time Shifted Viewing (TSV) is the term used for watching recorded programmes via VCR,

DVD or DVB:

Time shifted viewing on the actual day of broadcast (Day 0) is also called Vosdal

(‘Viewing on the same day as live’).

In the CIM TV study, time shifted viewing is identified as TSV up until 7 days after

broadcast (Day 0 to Day 7 inclusive). The total of this TSV is called “+7”

If a programme is viewed longer than seven days after broadcast, the viewing

process remains identified as VCR, DVD or DVB.

In order to determine whether the viewer is watching live or time-shifted, the content of the

broadcast on the viewed channel is compared with the broadcasts of the channel as it was

registered live on the so-called ‘Reference site’. This occurs in steps:

The viewing process is first compared to the so-called ‘Reference process’: the live

broadcast and all broadcasting to a maximum of 1.5 hours before that;

If no match is found, then a search is performed in the period up to seven days

before that for the first broadcast that does match;

If no match is found, the viewing process remains VCR, DVD or DVB.

Viewing processes cannot be ascribed to reference processes that are in the

future.

Time shifted viewing behaviour is only calculated for persons in the random sample of the

day of broadcast. Time shifted viewing behaviour of persons who do not belong to the panel

on the day of broadcast is reported as ‘Video’, ‘DVD’ or ‘HD’.

TSV will be available for the channels from the B-files, the channels reported per quarter (of

an hour) and the channels with Monthly Broadcasting Reports:

You can download the list of participating channels in the CIM TV study on the CIM website:

http://www.cim.be/sites/default/files/Media/Televisie/Documents/cimtv-stationlist.xlsx

As of 1 January 2016, CIM considers all ‘live’ television viewing + all time shifted viewing from Day 0 to Day 7 + Guest viewing as currency. From 01/01/2010 through 31/12/2015 the currency was ‘live+6+Guests’


2.2.2 Rating “live+7” calculation rules

a. Programmes

The rating of a programme is defined as the average percentage of viewers per second

within the reference universe of the broadcasting day that watched this programme live or

time shifted. The respondents in the random sample are weighted with a weight determined

by the extent to which they occur in the total population of viewers in the reference universe

of the broadcasting day. Time shifted viewers that were not present in the universe on the

broadcasting day are not taken into account.

%100

1

1

,,,,

R

r

r

R

r

tsvirrliveirr

i

w

pwpw

rating

ratingi :rating for programme i

R: number of respondents in the sample (broadcasting day)

wr,: weight of respondent r (broadcasting day)

p r,i, live: proportion of the programme i viewed live by respondent r

where the proportion is equivalent to:

i

ir,

, ir,duration

p

Vlive



p r,i,tsv : proportion of the programme i viewed time shifted by respondent r


i

ir,6

0

, ir,duration

,p

VTSV

TSV

tsv



TSV 0: time shift in the broadcasting day

TSV 1: time shift one day after the broadcasting day

...

TSV 7: time shift seven days after the broadcasting day


b. Timebands

%100

1

1

,,,,

R

r

r

R

r

tsvtrrlivetrr

t

w

pwpw

rating

ratingt : rating for time band t



p r,t, live: proportion of the time band t viewed live by respondent r

p r,t,tsv : proportion of the time band t viewed time shifted by respondent r

Averages per time band, over multiple days, should be calculated as follows: a data file is

calculated on a daily basis for the relevant target group including the rating per channel, per

second, rounded up to a minimum precision of six decimals. These data are then used to

calculate averages over multiple days, rounded off to the desired number of decimals.

2.2.3 “live+7” reach for a programme

The programme reach is defined as the percentage of viewers who watch one programme at

least once for a certain time, live or time shifted.

%

w

dw

reachR

r

r

R

r

I,rr

I 100

1

1

Jj

Jjd

Ir

IrIr

,

,,

0

1

I : series of programmes

dr,i: ‘dummy’ that indicates whether respondent falls within the programme rating according

to specific definition

J: criterion for the specific programme rating (e.g. 60 consecutive seconds)


wr: weight of respondent r (broadcasting day)

jr,I: the sum of live and time shifted viewing, calculated according to the specific conditions,

personal criterion of respondent for the programme series


2.2.4 “live+7” share

“Live+7” marketshare is calculated by dividing the “live+7” rating of a programme or time

band by the “live+7” rating of all channels in the relevant programme or time band.


application.

t,CH

tch,

tch

ratinglive

ratingliveemarketsharlive

"7"

"7""7" ,

with ch=channel, CH=all channels, t=time band

2.2.5 Specific Variables

a. Time Shift Delay

The Time Shift Delay (TSD) is the average number of minutes between the recording of

programmes and the playing of the programmes.

This is defined as the average number of minutes between the starting time of the Live

Broadcast Time (LBT) until six days before that and the start of viewing (Time Shifted

Viewing Moment = TSVM).

TSD = ( ∑(TSVMi – LBT0 )) / n

________

where 1 >= TSD =< 10080 minutes

TSVMi = Average Time Shifted Viewing Moment in day I (DD:HH:MM:SS in seconds)

i = 0,…, 6

LBT0 = Live Broadcast Time on day 0 (DD:HH:MM:SS in seconds).

n = number of TSV viewings

b. Market share in total TSV

This concerns the achieved TSV of a programme or time band with respect to the total TSV

(until six days after the broadcast).

This is a different result type than the Market Share TSV, because in the denominator the

sum of all the TSV occurring during a week is used.

Time shifted ratingt

MS TSV = (%) Time shifted rating (T)

t = 1,..., t series of Time Bands

T = 1+...+t total Time Band in days 1 to 7, where Time Bands do not overlap 0 =< MS TSV =< 100%


c. Share of TSV in rating

This is the share of TSV in the total rating of the programme or time band (Live plus TSV).

Rating TSV Programme/Time Band

TSV Share = (%) Total rating of Programme/Time Band

0 =< TSV Share =< 100%

d. ‘day share’ in TSV

This concerns the relative share of TSV according to programme or time band per day, as compared

to the total TSV for a programme or time band.

Time shifted rating programme or time band on day x

Day share in TSV = (%) Total Time shifted rating programme or time band

0 =< Day share in TSV =< 100%

2.2.6 Special calculation rules

a. Guests and Live+7

Only guests present on the actual day of broadcast (Day0) can be included in Live+7

analyses. Guest should therefore not be included in TSV other than Vosdal (‘Viewing on the

same day as live’).

b. Total Screen Usage and Live+7

With the introduction of TSV, the interpretation of Total Screen Usage (TSU) changed. Two

different approaches are possible: one based on the TSU for a particular broadcasting day

(TSULive+7) and one based on the TSU for a particular viewing day (TSUovernight).

TSUovernight = total amount of seconds on the ATA/ATB file

= Live + Vosdal + TSV from the past + OSU


Live: all stats with station-code2 ≠ VCR - Other - Play - DVB - Pay - DVD (ATA) and

VCR+DVD - PAY+DVB - Play+Other (ATB) and TSV-code3 = 0 ;

OSU: all stats with station-code = VCR - Other - Play - DVB - Pay - DVD (ATA) and

VCR+DVD - PAY+DVB - Play+Other (ATB) and TSV-code = 0 ;

What remains are stats with station-code = VCR - Other - Play - DVB - Pay - DVD

(ATA) and VCR+DVD - PAY+DVB - Play+Other (ATB) and TSV-code = 1 ;

These stats are equal to what is defined as Playback (see 2.4 Calculation rules for

“Live+Playback) namely the total of Vosdal and TSV-1 + TSV-2,… ,TSV-7 = TSV from

the past.

TSULive+7 = Live + Vosdal + TSV+1->+7 + OSU

Live: identical to Live as defined for TSUovernight ;

Vosdal: Viewing On the Same Day As Live (delivered in the TSA/TSB files) ;

OSU: identical to OSU as defined for TSUovernight ;

TSV+1->+7: TSV is delivered in the TSA/TSB files in the days after the broadcasting

day.

The scheme below summarizes the difference between TSUovernight and TSULive+7.

Broadcasting day x

Viewing day x+0

Viewing day x+1

… Viewing day

x+7

Live day x Live day x Live day x … Live day x

Vosdal day x Vosdal day x Vosdal day x … Vosdal day x

OSU day x OSU day x OSU day x … OSU day x

TSV x-1 =>-7

TSV x+1 …

TSV x+1 =>+7

= = = = =

TSUovernight TSUlive+0 TSUlive+1 … TSUlive+7

TSUlive+1 may not be used in analyses of programme variables because this value is

intermediate and not complete.

2 Not to be confused with the Source code, a new field compared to the former AUA-AUB file that indicates which device generates the signal (see 5.3 Description ATA-ATB and TSV files). 3 A new field that has been added to the ATA-ATB file which indicates if a statement is time shifted yes or no (0=no, 1=yes).


c. Sources

The introduction of the Source code (field 9, see 5.3 Description of ATA-ATB and TSV files)

makes it possible to analyze the strength of different sorts of peripheral equipment4.

The scheme below illustrates the repartition of the different sources, how this evolves over

time and how this results in TTV and TSU (fictive example based on seconds). In the

columns you’ll find the different sources and the results for TTV and TSU. In the rows you’ll

find the results for Live, Live+Vosdal, Live+1,…, and the OSU (based on the Station code).

When moving from Live to Live+Vosdal, Live+1,… the amount of seconds per source will

increase. The only exceptions are TV and Satellite as these devices do not allow recording a

program and hence it is impossible to have TSV on these sources.

4 The calculation of Total TV (TTV) and Total Screen Usage (TSU) (as described in 2.1.4 Rating calculation rules) remains unchanged and is still based on the Station code.

Source Code TV=0 VCR=1 SAT=2 Pay=3 DVD=4 PLAY=5 DVB=6 Other=7 CTV=8Total Source

0 => 8

TOTAL TV

2006+

Total

Screen

Usage

Live 34.496 1.411 546 0 17.495 381 234.041 48 579 288.998 288.998 366.506

Live+0 (Live+Vosdal) 34.496 1.446 546 0 18.524 387 246.815 48 579 302.843 302.843 351.596

Live+1 34.496 1.486 546 0 18.917 387 252.540 48 579 309.000 309.000 357.753

Live+2 34.496 1.509 546 0 19.075 387 255.069 48 579 311.710 311.710 360.463

Live+3 34.496 1.568 546 0 19.206 387 256.632 48 579 313.464 313.464 362.217

Live+4 34.496 1.684 546 0 19.299 388 257.701 48 579 314.741 314.741 363.494

Live+5 34.496 1.757 546 0 19.353 388 258.589 48 579 315.756 315.756 364.509

Live+6 34.496 1.846 546 0 19.437 388 259.376 48 579 316.716 316.716 365.469

Live+7 34.496 1.875 546 0 19.499 388 260.103 48 579 317.535 317.535 366.288

OSU 0 1.248 0 0 14.184 3.200 27.738 619 1.764 48.753 48.753

TSU/Source 34.496 3.124 546 0 33.683 3.587 287.841 667 2.343 366.288 366.288 366.288


2.3 Algorithm for the evaluation of advertising campaigns (“Live+7”)

2.3.1 General calculation rules for campaign evaluations

The general calculation rules for the CIM TV study have been used since 1997.

a. Main variables

The main variables in the post-evaluation of an advertising campaign are:

Total coverage in %: percentage of target universe having seen a minimum of one

spot

OTS: Opportunity To See. This is the average number of contacts each person has

with the spot.

Contact distribution: cov%(k): percentage persons in the target universe who had k

contacts with the spot.

Daily GRP of each spot

Total GRP of the campaign

Like all TV variables, all computations described in this chapter can be computed for a

particular target. The only difference between computation for one target and for the

Universe is that the sample is restricted to the persons belonging to the specified target

group. We will therefore explain the algorithm for the Universe only.

b. Definition of a campaign

A campaign is a list of spots (events) broadcasted during a limited period of time.

In CIM TV a campaign may not cross year boundaries.

It is never a good solution to evaluate campaigns over too long periods because this

will reduce the quality of the panel: not only some audimeters may not answer on

specific days but also more members may enter and exit from the panel during the

period.

A campaign is defined in the following way:

Day id Spot id Date Channel Start time Duration

d1 s1 18-aug-2008 CH3 19:00:00 31

d1 s2 18-aug-2008 CH4 20:00:00 27

d2 s3 19-aug-2008 CH2 19:30:00 45

d2 s4 19-aug-2008 CH1 17:05:13 71

d3 s5 22-aug-2008 CH1 25:12:17 11

d4 s6 15-sep-2008 CH6 23:47:58 5

d5 s7 16-sep-2008 CH5 19:45:34 15

D: number of days where one spot of the campaign is broadcast, in this example D=5.

S : number of spots, in example, S=7


c. Panel constitution for a specific campaign

To compute the post-evaluation, the algorithm will first define a campaign panel.

In the CIM TV study the panel used is a Union panel: all participants of the Audimetrie panel

on one or more of the campaign days are a member of the panel.

In the example above, all members of the Audimetrie panel present on the 18th or 19th or

22nd of August or the 15th or 17th of September will be included in the campaign panel.

Earlier (see 2.1.1) the use of the daily weights for each individual panel member was

described. However, at the campaign level, independent of the daily weights, a person

present all days of a campaign will have a higher importance than a person present only a

small number of days. In order to reflect this difference, each person in the campaign panel

receives his own “campaign weight”.

Let’s define:

dW(d,m): “daily weight” of member m for day d (d is the day of broadcast)

cW(m): “campaign weight” of the member (a constant for the campaign)

with

m: one member

M: the number of members

The example below helps to understand how the campaign weights are computed.

daily weights of m1 daily weights of m2 daily weights of m3

d1 100 80 70

d2 0(Absent) 70 72

d3 90 0(Absent) 68

d4 95 95 74

d5 105 105 76

The constant weight is computed as the sum of daily weights (0 if absent) divided by D, the

total number of days of the campaign

cW(m1) =(100+90+95+105)/5 = 78

cW(m2)=(80+70+95+105)/5=70

cW(m3)=(70+72+68+74+76)/5=72

d. GRP’s for an individual spot

The GRP’s published at the spot level are computed with the member daily weights.

So we can define the GRP’s for spots as:

M

m

dm

M

m

dmsm

s

dW

dWc

GRP

1

),(

1

),(),( *

% *100,0


with c(m,s)=1 (true), if nS(m,s) >= ceiling(l(s)/2)

c(m,s)=0 (false), if nS(m,s) < ceiling(l(s)/2)

and l(s) = the length of a spot s in seconds

nS(m,s) = the number of seconds of spot s viewed by member m, consecutive or not.

e. GRP’s, coverage and OTS for a campaign

The total GRP for a campaign = GRPT%.

The total GRP of the campaign can be defined as the sum of GRP’s for the individual spots in

a campaign:

S

s

sGRPGRPT1

%%

Similar to the notion of contact of a member with one spot, we can define the contact of a

member with the whole campaign as the sum of contacts with individual spots:

S

s

smcmcT1

),()(

The total coverage can then be expressed as:

M

m

M

m

mcW

mcWv

COVT

1

1

)(

)(*

% *100,0

With

v=1 if cT(m) > 0 (contact with at least one spot in the campaign),

v=0 if cT(m) = 0 (not a single contact with one of the spots in the campaign).

Logically, the average OTS can then be expressed as:

Average %

%

COVT

GRPTOTS

f. Computing the contact distribution for a campaign

Computing the contact distribution for a whole campaign, cannot be based simply on data

for individual spots. These data involve daily weights that can differ from day to day per

member and don’t take into account the difference between members in campaign weight.

It is possible to compute the total GRP based on campaign weights per campaign panel

member:

M

m

S

s

M

m

mcW

mcWsmc

cGRPT

1

11

)(

)(*),(

% *100,0


We can now define the % contacts for any given k (k being a number of contacts):

M

m

M

m

mcW

mcWv

kCOVTk

1

1

)(

)(*

)%( *100,0

With

v=1 if cT(m)=k,

v=0 if cT(m) ≠ k.

However, neither the cGRPT% nor the COVTk%(k) may ever be published:

Given the fact that GRPT% is based on daily weights but cGRPT% is based on

campaign weights, often GRPT% will not be equal to cGRPT%.

Since COVTk%(k) is also based on campaign weights it cannot be published.

One of the priorities of the CIM TV algorithm is to allow the publication of GRP’s for

individual spots during the campaign without risk for modification of these values

afterwards. Therefore, only the GRPT% can be published.

To calculate a publishable contact distribution, a correction algorithm is used that transforms

the contact distribution computed with campaign weigths in such a manner that it respects

following mandatory conditions:

GRPT% must be unchanged, it is the sum of the GRP% of each spot;

COVT% defined above must also remain unchanged;

OTS must evidently remain GRPT%/COVT%.

Since the objective is to establish following formula,

K

k

kkCOVTkGRPT1

*)%(% ,

the values of COVTk%(k) will have to be changed to keep the previous equation true

while maintaining also the following relation:

K

k

kCOVTkCOVT1

)%(%

The correction algorithm seeks to establish in five steps for each number of contacts a transformation value so that:

1.

K

k

kk kdGRPT1

).(% or

K

k

k cGRPTGRPTk1

%%. , and

2. )(%1

k

K

k

kdCOV

or

K

k

k

1

0


Step 1.

Knowing that the average%

%

COVT

GRPTOTS , we define )int( averageOTSF .

This value is used as a pivot value. It allows to compute the % of members (coverage) in

the campaign panel present on each side of the pivot value in the contact distribution:

On the lower end:

F

k

m kCOVTkD1

)%(

On the higher end:

K

Fk

M kCOVTkD1

)%(

Step 2.

β is defined as the coverage for any given OTS level divided by the total coverage on its side

of the pivot value, with a negative sign if it falls in the lover end of the contact distribution:

m

kD

kCOVTk )%( for Fk

M

kD

kCOVTk )%( for Fk

with

K

kk

1

0

Step 3.

Multiply for each OTS level the β-value with the number of OTS.

A is defined as the sum of these multiplications: k.AK

kk

1

Step 4.

This allows calculating n,...,, 21 from the equation

kkA

cGRPTGRPT .

%%

taking 1) and 2) into account.

Step 5.

Finally, the values of COVTk%(k) can be adapted for all k as follows:

kkCOVTkkCOVTk )%()%(

Final check: Check once more that coverage and OTS are unchanged and that following

formula is true

K

k

kkCOVTkGRPT1

*)%(%


Hereunder, you can find an example of a correction by the algorithm.

Campaign evaluation results Total of daily GRP (sum of GRP's/event) 250 Coverage % whole "campaign" 68 Average OTS 3,68

Step 0 Step 2 Step 3 Step 4

Number of contacts k

cov%(k) Campaign GRPs k*cov%(k) (not

published) Beta(k) Beta(k)*k Alpha(k)

Corrected cov%(k)

Corrected GRPS(k)

1 10,00 10,00 -0,26 -0,26 3,30 13,30 13,30 2 8,00 16,00 campaign GRP = 319,00 -0,21 -0,42 2,64 10,64 21,28 3 20,00 60,00 F = 3,00 -0,53 -1,58 6,60 26,60 79,80 5 8,00 40,00 0,27 1,33 -3,34 4,66 23,28 6 5,00 30,00 Dm = 38,00 0,17 1,00 -2,09 2,91 17,46 7 1,00 7,00 DM = 30,00 0,03 0,23 -0,42 0,58 4,07 8 7,00 56,00 0,23 1,87 -2,93 4,07 32,60 9 0,00 0,00 0,00 0,00 0,00 0,00 0,00 10 5,00 50,00 0,17 1,67 -2,09 2,91 29,10 11 0,00 0,00 0,00 0,00 0,00 0,00 0,00 12 3,00 36,00 0,10 1,20 -1,25 1,75 20,96 13 0,00 0,00 0,00 0,00 0,00 0,00 0,00 14 1,00 14,00 0,03 0,47 -0,42 0,58 8,15

Total 68,00 319,00 0,00 5,50 68,00 250,00

campaign GRP A

COVTOT

GRPTOT

Double check Recomputed COVTOT 68 Difference with COVTOT should be 0 True

Recomputed GRP TOT 250 Difference with daily GRPTOT should be 0 True

Input data

Pivot value +

Split coverage%

Step 1

Intermediary values Result

Step 5

See also excel file “Campaign Evaluation Algorithm”


2.3.2 Adaptations to handle time shift viewing in the post-evaluation algorithm

Time Shifted Viewing can create a slightly different situation:

c(m,s) was always 0 or 1.

It now becomes the total number of Live and Time Shifted contacts of member m with

spot s.

Its value can be 0, 1 or higher.

If we define

LENGTH: the number of seconds of spot s

LIMIT : ceiling (l(s)/2,0)

NS: total number of seconds of this spot seen by member,

eventually partly live, partly in TSV

the value of c(m,s) is defined in the following way

0 if NS < LIMIT

1 if NS between LIMIT and (LENGTH + LIMIT -1)

2 if NS between (LENGTH + LIMIT) and (2* LENGTH + LIMIT -1)

N if NS between ((N-1) * LENGTH + LIMIT) and (N* LENGTH + LIMIT -1)

Another modification is the number of values in the distribution contacts.

With TSV, the distribution contacts can contain more values than the actual number of

spots.


2.4 Calculation rules for “Live+Playback”

2.4.1 Basic Principle

Live+Playback is a new “viewer centric” measure. Whereas Live and Live+7 are “broadcast

centric” measures. The starting point for Live+Playback is the viewer who watched a certain

channel on a specific day independently of the broadcasting day of the programme(s) or

time band(s) involved. In other words: both live and time shifted viewing on the specific day

are taken into account.

The time of viewing is 1 specific day only. The content of the viewed broadcasts covers a

maximum of 7 days (day of viewing and the broadcasts of six days before).

Live+Playback may only be reported for Total TV and up to channel level.

Live+Playback cannot be calculated for programmes because programmes may be

broadcasted in the past and the ‘live’ audience is not taken into account.

Live+Playback on one specific day is the total of:

a/ Live viewing on a that specific day (Live)

b/ The Time shifted viewing on the actual day of viewing (TSV 0)

c/ The Time shifted viewing of programmes broadcasted until 7 days before that specific

day (TSV -1 until TSV -7).

The total of b/ and c/ is called “Playback”

Day -7 Day -6 Day -5 Day -4 Day -3 Day -2 Day -1 Day 0Station A

Broadcast 1 TSV -7Station A

broadcast 2 TSV -6Station A







broadcast 9

Station A

broadcast 10

Station A

broadcast 11

Station A

Viewing DayBroadcasting day

Pla

ybac

kLi

ve

Live

+ P

layb

ack


2.4.2 Rating “Live+Playback” calculation rules

a. One a specific day

The rating on a specific day is defined as the average percentage of viewers per second

within the reference universe of this viewing day that watched live or in playback. The

respondents in the random sample are weighted with a weight determined by the extent to

which they occur in the total population of viewers in the reference universe of the viewing

day. All viewers that are present in the universe on the viewing day are taken into account.

%100

1

1

,,,,

R

r

r

R

r

playbackdrrlivedrr

d

w

pwpw

rating

Ratingd :rating for one specific day, the viewing day

R: number of respondents in the sample on the viewing day

wr,: weight of respondent r on the viewing day

p r,d, live: proportion of the day (d) viewed live by respondent r

p r,d, playback: proportion of the day (d) viewed in playback by respondent r


d

d

lived

V

duration

p

r,

, r,

V r,d: Viewing duration of respondent r for the specific day (d)

Duration d: duration of the day (24h)

d

dTSV

TSV

playbackd

V

duration

p

r,6

0

, r,

TSV 0: time shifted viewing on the viewing day

TSV-1: time shifted viewing from content broadcasted 1 day before the viewing day

TSV-2: time shifted viewing from content broadcasted 2 days before the viewing day

...

TSV -7: time shifted viewing from content broadcasted 7 days before the viewing day

Rating can only be calculated for a (number of) Channel(s), namely a channel with TSV

data, see station list provided by the CIM.


2.4.3 Reach “Live+Playback” calculation rules

The channel reach for one specific day is defined as the percentage of viewers who watch

that channel at least once for a certain time during a specific day, live or in playback.

%100

1

1

,

R

r

r

R

r

chrr

ch

w

dw

reach

Jj

Jjd

chr

chrchr

,

,,

0

1

ch : channel

dr,ch: ‘dummy’ that indicates whether respondent falls within the channel rating according to

specific definition (viewing threshold)

J: criterion for the specific channel rating (e.g. 60 consecutive seconds)


wr: weight of respondent r (viewing day)

jr,ch: the sum of live and playback viewing, calculated according to the specific conditions,

personal criterion of respondent for the channel.

2.4.4 “Live+Playback” share

”Live+Playback” share is calculated by dividing the ratingd of a specific channel by the

ratingd of all channels on the specific day.


application.

d,CH

dch

dch

rating

ratingemarketsharplaybacklive

,

,""

with ch=channel, CH=all channels, d=day


3. Reporting rules

3.1 Live, Live+7 and Live+Playback rating/reach

Given the introduction of Time Shifted Viewing, it has become possible to make different

calculations for rating and reach:

Live rating/reach : viewing during the live broadcast,

Live+7 rating/reach : viewed during the live broadcast or until 7 days later;

Live+Playback rating/reach : viewed live and playback (broadcasted until 7 days

before the viewing day).

Live and Live+7 rating and reach emit from 1 broadcast and when it is viewed, live or later:

Live + Playback rating and reach proceed from the viewers who, on 1 day, view one channel

live or broadcasts from the near past of one channel in playback.

Schematically, the distinction between Live, Live+7 and Live+Playback looks like this.

Day -7 Day -6 Day -5 Day -4 Day -3 Day -2 Day -1 Day 0 Day 0 Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7

Station A

Broadcast 1 TSV -7Station A

broadcast 2 TSV -6

Station A

broadcast 3 TSV -5

Station A





broadcast 8 TSV -0

LiveStation A

broadcast 9

Station A

broadcast 12

Station A

TSV of 12

Station A

TSV of 9

LiveStation A

broadcast 10

Station A

broadcast 13

Station A

TSV of 3

Station A

TSV of 8

LiveStation A

broadcast 11

Station A

broadcast 14

Station A

TSV of 7

Station A

TSV of 10

Station A

TSV of 5

Live Live TSV 0 TSV 1 TSV 2 TSV 3 TSV 4 TSV 5 TSV 6 TSV 7

Viewing day

6

Broadcasting day One Specific Day

Station A

Playback


3.2 Live, Live+7 and Live+Playback

The calculation of the share of a broadcast or time band in total viewing becomes more

difficult, due to the introduction of Time Shifted Viewing. Depending on the jumping-off

point, both the viewing volume of a broadcast and the total viewing volume can be defined

differently.

The software applications must enable two different bases for analysis of the data:

The broadcast: the content as it is broadcast live on day X.

The viewer: the period when a viewer watches content, live on day 0 or time shifted.

The term ‘content’ broadly encompasses all types of broadcasts here: programmes,

programme components, spots, etc.

The CIM has decided that for ‘content’ only 3 results may be published:

Live share (based on a broadcast)

Live+6 share (based on a broadcast between 01/01/2010 and 31/12/2015)

Live+7 share (based on a broadcast since 01/01/2016).

And for only time combined with channel:

Live+Playback share (based on the viewer).

Any other analysis may only be used for internal purposes.

3.2.1 Live share

Until the introduction of data regarding Time Shifted Viewing, the market share was

calculated as the percentage of viewers who watched a broadcast (live by definition) or time

band, divided by the total viewing audience (see 2.1.4 above).

This notion remains in existence by the name ‘Live share’.

3.2.2 Live+6 share

By measuring Time Shifted Viewing, it became obviously interesting to add up the sum of

live and time shifted viewing (on the day itself until six days later) for a broadcast and to

compare it with the broadcasts on other channels with the same start and end time.

This variable is called ‘Live+6 share’ and may only be reported for the period 01/01/2010 -

31/12/2015


3.2.3 Live+7 share

The variable, ‘Live+7 share’ has the same characteristics as Live+6 share but include one

additional day of time shifted viewing. All reporting of programmes, broadcast schedules, as

well as spots, blocks of advertising and time bands that were broadcasted since 1st of

January 2016, should be done using this variable.

Example:

The programme ‘Thuis’ was broadcast on 01/02/2016 and was viewed on that day as well as

later. This programme was then (re)watched later and distributed up to 7 days after broadcast.

Until 08/02/2016, time shifted viewing of ‘Thuis’ was allocated to 01/02/2016.

The ‘Live share’ of Thuis corresponds with the market share as was used until the

introduction of TSV: the percentage of viewers of ‘Thuis’ in relation to the total ‘live’ viewing

audience.

The ‘Live+7 share’ is the percentage of viewers of ‘Thuis’ in relation to the total viewing

audience. The total viewing audience are the viewers of Thuis and all time slots of

broadcasts that were programmed parallel on 1/9/2008 (from 9:15 pm – 10:05 pm) and

that were viewed both live (on 01/02/2016) and time-shifted (01/02/2016 (VOSDAL)-

08/02/2016).

This type of reporting can also be applied to a channel's broadcast schedule.

For example: RTL-TVI between 8 pm and 10:30 pm on 01/02/2016.

All viewers (both live and time shifted until and including 08/02/2016) of the programmes

within this time span on Canvas and the programmes on the channels that are programmed

parallel belong to the total Live+7 viewing audience.

This idea applies to programmes, broadcast schedules, as well as spots, blocks of

advertising and time bands.

3.2.4 Live+Playback share

The starting point here is the viewer who watched a certain channel on a specific day, separate

from the fact of whether the broadcast was followed live or dated from an earlier time (until

six days ago).

The time of viewing is thus one specific day only. The content of the viewed broadcasts covers

a maximum of 8 days (day of viewing until and including seven days before).

The share of a channel is defined here as the ‘Live+Playback share’.

To keep the user from becoming lost in all possible combinations of viewing times and

programmes, time bands, spots and so on, reach, rating and share may only be reported for

Total TV and up to channel level.

3.2.5 “+7” and “Playback”

In the CIM TV study, time shifted viewing is identified as TSV up until 7 days after broadcast

(Day 0 to Day 7 inclusive). The total of this TSV is called “+7”


The total of the time shifted viewing on the actual day of viewing (TSV 0) and the time shifted

viewing of programmes broadcasted until 7 days before that specific day (TSV -1 until TSV -

7) is called “Playback”.

3.2.6 Monthly figures including Time Shifted Viewing

Calculations of reach and rating for a month may only be reported on the basis of a channel.

In a month of 30 days, the figures must be calculated based on what was watched from day

1 to 30, inclusive.

So this is not about the programming of those days that can also be viewed seven days after

the month has ended.

3.2.7 Campaign evaluation

Since the currency is ‘Live+7+guests’, the balance of advertising campaigns may only be

published on the basis of Live+7+guests.

In practical terms, this means:

Day 1 is the day of the broadcast of spot 1,

The last campaign day is the day of broadcast of the last spot + 7 days.


3.3 Source citation

Every report without exception must state the conditions or parameters that resulted in the

specific reported channel reach.

As indicated above, CIM TV considers ‘Live+7+guests’ as the currency.

Every document and presentation must state the source of the data: ‘Source: CIM TV’ as

well as the target group, hour slot and all additional parameters that determined the

universe of the analysis. Also the company who generated the data needs to be mentioned.

The fact that a channel has a detailed listing in AUB, has a monthly broadcasting report or

just per quarter must be considered.

If it does not involve complete data, this must be clearly indicated:

If Live only: ‘Source: CIM TV – Incomplete figures, only live viewing’

If Time Shifted only: ‘Source: CIM TV – Incomplete figures, only time shifted viewing’

3.4 Rounding

The CIM recommends rounding to one (1) digit after the decimal point for final reporting.

An exception is the export of data, in databases or in other software, where new calculations

can be made. Greater precision remains necessary there (at least 4 digits after the decimal

point).


4. Available files An Excel file with a detailed description of ATA-ATB and TSV files is available upon request at

CIM: 02/661 31 53 or [email protected]. http://www.cim.be/nl/televisie/methodologie

http://www.cim.be/fr/television/methodologie

Field 10 TSV is added to the TSV part in the ATA and ATB file.

4.2.2 TSV file

The TSV file includes the recording time and the re-viewing time.

The following variables are included in the file:

Recording date

Family number

Channel code

Time of broadcast

Viewing duration

Date of re-viewing

Time of re-viewing

Source code

4.3 Top 20 and Top 100

A weekly Top 20 of the most watched TV programmes in the North and the South is

available in the public section on the CIM website:

This top 20 contains the rating and market share of the best viewed programmes for a

week. Data are published on the TSV+7+Guest currency and are therefore available 8 days

after the weeks end.

The annual Top 100 with the 100 most viewed programs is also available based on the

TSV+7+Guest currency.. It is available by the end of January of each year.

Field Description Start.pos End.pos Length

9 Source code 44 45 2

10 Time Shifted Viewing 46 46 1

mailto:[email protected]

http://www.cim.be/nl/televisie/methodologie

http://www.cim.be/fr/television/methodologie